  |
 |
 |
 |
 |
|||
 |
|||||||
 |
|
|||||||||||||||
HuGE Review
|
|
This paper was published with modifications in Genet Med 1999 Nov-Dec; (7):332-9 Medium chain acyl-coA dehydrogenase deficiency (MCADD) |
by Sophia S. Wang1, Paul M. Fernhoff2, W. Harry Hannon3, and Muin J. Khoury1
1Centers for Disease Control and Prevention, National Center for Environmental Health, Office of Genomics and Disease Prevention, Atlanta, GA
2Emory University, Department of Pediatrics, Division of Medical Genetics, Atlanta, GA
3Centers for Disease Control and Prevention, National Center for Environmental Health, Environmental Health Laboratory, Atlanta, GA
March 3, 1999 (Updated May 26, 1999)
Internet Sites Medium chain acyl-CoA dehydrogenase (MCAD) is a tetrameric flavoprotein essential for the beta-oxidation of medium chain fatty acids. MCAD deficiency (MCADD) is an inherited error of fatty acid metabolism. The gene for MCAD is located on chromosome one. One variant of the MCAD gene, G985A, a point mutation causing a change from lysine to glutamate at position 304 (K304) in the mature MCAD gene has been found in 90% of the alleles in MCADD patients identified retrospectively. Although this review focuses on the K304 (G985A) MCAD variant, data regarding disease outcomes are not specific to this allelic variant. There is a high frequency of MCADD among people of Northern European descent, which is believed to be due to a founder effect. Because MCADD is inherited in an autosomal recessive manner, it can occur only if a person has two copies of the G985A variant, one copy of the G985A variant plus one copy of another MCAD mutation, or two copies of other MCAD mutations. People who are heterozygotes with one normal allele do not express symptoms of MCADD. Of patients clinically diagnosed with MCADD, 81% who have been identified retrospectively are homozygous for G985, and 18% are compound heterozygotes who have G985 as one of the disease alleles. Clinical data on the probability of clinical disease indicates that MCADD patients are at risk for the following outcomes: hypoglycemia, vomiting, lethargy, encephalopathy, respiratory arrest, hepatomegaly, seizures, apnea, cardiac arrest, coma, and sudden and unexpected death. Long-term outcomes include developmental and behavioral disability, chronic muscle weakness, failure to thrive, cerebral palsy, and attention deficit disorder (ADD). It is unknown whether differences in clinical disease specific to allelic variants exist. Risk factors that may modify disease severity are age when the first episode occurred, fasting, presence of infection, and family history of MCADD. Acute attacks must be treated immediately with large intravenous doses of glucose. For those diagnosed, long-term management of the disease includes preventing stress caused by fasting and maintaining a high-carbohydrate, reduced-fat diet. Hospitalization costs attributable to morbidity and mortality from MCADD are unknown; MCADD is not a diagnosis in the International Classification of Disease, 10th Revision (ICD-10) codebook. There appears to be a substantial number of asymptomatic MCADD individuals. Unfortunately, it is not known how to distinguish which individuals will manifest symptoms and which individuals will remain asymptomatic. Several tests are available to detect MCADD, including DNA-based tests for G985 mutations, which usually involve the use of the polymerase chain reaction (PCR). The main technique, which detects abnormal metabolites in blood or urine, consists of tandem mass spectrometry (MS/MS). State programs are beginning to offer screening in newborns for MCADD using (MS/MS). In addition, Neo Gen Screening (Pittsburgh, PA 15220) currently offers voluntary supplemental newborn screening for MCADD to birthing centers. While MCADD satisfies most of the requirements for newborn screening, it does not satisfy the first criterion, that the natural history of the disease and its clinical outcomes are clearly understood. Indexing terms: medium-chain acyl-coenzyme A dehydrogenase The medium chain acyl-coA dehydrogenase (MCAD) gene is located at chromosome 1p31. The protein formed, MCAD, is an oxidoreductase enzyme that catalyzes the first step of the medium-chain fatty acid oxidation cycle, a two-electron oxidation of fatty acyl-coA thiolesters. MCAD is part of the acyl-coA dehydrogenase (flavoprotein) family and is one of five acyl-coA dehydrogenases. MCAD is one of several acyl co-A dehydrogenases that catalyze the beta-oxidation of fatty acids. The MCAD enzyme is a homotetramer, composed of two polypeptide chains. While 26 MCAD gene variants have been reported (Table 2) (1), only nine of these MCAD variants are documented in OMIM (2). This review will focus on the LYS329GLU (G985A) variant, the MCAD mutation reportedly found in 90% of all retrospectively identified MCAD deficient patients' alleles, and for whom 81% of all MCAD patients are homozygous; 18% of MCAD patients are compound heterozygous for G985A. Caucasian individuals of Northern European descent exhibit the highest frequency of MCADD genotypes. The carrier frequency of G985A among this group is 1:40-1:100 and the homozygote frequency is 1/6,500-20,000 (3). Haplotype analyses have suggested that the G985A MCAD mutation is a founder effect from a single person in a Germanic tribe (4). Table 1 lists the frequencies of the G985A allelic variant for the MCAD gene by country. Only two studies have been conducted on a population-based prospective cohort of newborns; most others are retrospective studies. While heterozygotes have been identified with this method, few homozygotes have been identified; the frequencies of homozygosity shown in Table 1 are therefore mostly estimates based on the frequency of heterozygosity and using Hardy-Weinberg equilibrium. Despite this limitation, Ziadeh and colleagues' (1995) population-based study of U.S. newborns demonstrated a frequency of homozygosity (1/20,408) that is similar to those estimated in some northern European countries (5). Ziadeh and colleagues studied the first 80,371 newborns from a supplemental newborn screening program in Pennsylvania and eastern Ohio for over 35 metabolic disorders (5). Despite the heterogeneity of the population, which included high proportions of Eastern Europeans and Blacks, nine MCADD newborns were diagnosed from the first 80,371 newborns screened between November 1, 1992 and September 30, 1994 by tandem mass spectrometry. However, upon DNA sequencing of the nine MCADD newborns, the distribution of the G985A allele was found to be different than previously described. Among MCADD individuals, 44% (four of nine) were homozygous for G985A and 56% (five of nine) were compound heterozygotes, versus the 81% homozygous and 18% compound heterozygotes previously calculated based on retrospective data. In the U.K., however, Pollitt and Leonard's (1998) prospective study of MCADD through the British Paediatric Surveillance Unit identified 57 people in England with MCADD, for an MCADD incidence of 4.5/100,000. Of the 45 people with MCADD whose DNA has been analyzed thus far, 36 of 45 (80%) were homozygous for G985A, and 9 of 45 (20%) were heterozygous for G985A (6), which is consistent with data from retrospective studies (since data are still pending, this study is not listed in Table 1). It is important to note that most of the retrospective studies are conducted using spare blood samples and their results are not necessarily representative of the population, nor can they accurately assess incidence of the disorder. Only from population-based prospective cohort studies can incidence of a disorder be accurately assessed. Research on MCADD is in its infancy, with most of the published articles having concentrated on the biochemical and molecular aspects of the disease. The natural history and long-term outcomes of patients with the disorder are still not clearly defined, nor is determining who will present with disease and who will remain asymptomatic. Clinical outcomes in relation to specific allelic variants are also not clearly defined. Population and epidemiologic information gathered from pilot and supplemental testing programs currently being conducted will begin to provide much needed information about the clinical spectrum of the disorder. MCAD Deficiency (MCADD) The first MCADD patients were identified in 1982, and between 1982 and 1994, more than 200 MCADD patients were identified (23). MCADD is an autosomal recessive genetic disorder. MCADD occurs only when there are mutations in both MCAD alleles, leading to an abnormal protein product and therefore inefficient enzymatic activity to metabolize medium chain fatty acids. The G985A MCAD variant occurs when glutamate is substituted for lysine at position 304 in the mature MCAD gene. Carriers who have one copy of a variant MCAD allele and one copy of the normal MCAD gene, have adequate enzyme levels. Individuals who are homozygous or compound heterozygous for an MCAD mutation may exhibit some clinical manifestations of MCADD. Differences in clinical manifestations specific to allelic variants are not yet defined. The incidence of MCADD can only be calculated from prospective population-based studies. In Ziadeh and colleagues' (1995) study conducted in Pennsylvania and eastern Ohio, nine newborns with MCADD were identified out of 80,371 tested, via tandem mass spectrometry for an incidence rate of one in 8,930 (5). Updated data from the same population has identified 16 newborns with MCADD out of 211,067 tested, for an incidence rate of one per 13,192 (24). These rates are higher than the MCADD incidence rate found through Pollitt and Leonard's (1998) prospective study (one per 22,222) in the U.K. (6). Although the frequency of the allele is rather high in Caucasian populations, based on rates of heterozygosity and under Hardy-Weinberg conditions, there is a greater number of expected MCADD cases than are currently diagnosed. On the basis of the heterozygote frequency in New South Wales, Wilcken and colleagues (1994) estimated that the MCADD incidence rate should be 1/15,000 to 1/20,000 or 42 cases in their 1975-84 study (25). However, only 9 MCADD individuals (21% of expected) were identified, and only 7 of an expected 31 from 1985-91 were identified. Similarly, Fromenty and colleagues (1996) found that of 414 French blood donors, the mutant gene frequency of 1/138 produces a homozygote frequency of 1/19,000 according to Hardy-Weinberg, for an expected 42 MCADD individuals (14). However, only six cases of MCADD were reported. Similar discrepancies have been seen elsewhere. The numbers of MCADD cases observed have been consistently lower than the number expected, suggesting a penetrance of less than 100%. Two hypotheses have been offered to explain this lower than expected number of cases. First, symptomatic newborns may be misdiagnosed as having SIDS or Reye's or die before MCADD is detected. Second, a substantial number of people with MCADD may be asymptomatic and unidentified. Although misdiagnosis and early deaths may account for some unidentified cases, they are not likely to account for all of them. There is a wide spectrum of disease severity ranging from no symptoms to death. The following two sections summarize the disease outcomes associated with MCADD, and the risk factors for presenting with the various outcomes. It is unknown whether disease severity is associated with specific allelic variants. The following sections are therefore based on MCADD diagnosed clinically or biochemically, and are not outcomes specific to the G985A allelic variant. Clinical signs and symptoms usually first occur during a child's first three years of life (average is one year). Neonatal onset is possible, and presentation rarely occurs in adults. There is no typical presentation of MCADD (26). The most common features include vomiting and lethargy. Hypoglycemia, encephalopathy, respiratory arrest, hepatomegaly, seizures, coma, apnea, cardiac arrest, and sudden death have also been documented. Long-term outcomes may include developmental and/or behavioral disabilities including mental retardation, cerebral palsy, and ADD. Table 3 displays results from three studies that have documented symptoms observed in clinically diagnosed MCADD patients. In summary, although the signs and symptoms of MCADD are variable, they are similar to those seen in other encephalopathies: vomiting, lethargy, seizures, and progressive coma. Routine clinical laboratory findings include hypoketotic hypoglycemia, with hyperammonemia and abnormal liver function. Virtually all information published on clinical outcomes related to MCADD is in the form of case studies. Systematic evaluations of morbidity and mortality in a population-based study have not been conducted. Furthermore, since MCADD is not listed in the ICD-10 codebook, estimates of morbidity, hospitalization burden, and annual costs of the disease cannot be calculated. Given the increasing demand for newborn screening of MCADD, there is a critical need for long-term follow-up of people with MCADD in order to elucidate the natural history of the disease. Clinical symptoms and outcomes at presentation Numerous case reports and case series have documented clinical outcomes for people with MCADD who were untreated prior to presentation. Few studies, however, have reviewed the outcomes of MCADD individuals beyond the scope of the case reports. Two studies that did attempt to quantify the number of people affected or presenting with various clinical outcomes are summarized in Table 3. To date, Iafolla and colleagues (1994) have reviewed the largest number (120) of MCADD cases. All of the patients in the study were of European descent: 112 from the United States, and the remaining 8 from the United Kingdom, Australia, Canada, or Ireland (27). These cases were clinically identified through physicians and subsequently analyzed and confirmed by at least two methods of MS. Touma and Charpentier (1992) reviewed outcomes for 65 MCADD patients identified upon acute onset of the disease at hospitals in France and who had their MCADD diagnosis confirmed by GC/MS (28). A third study summarized the outcomes of the first 23 reported cases of MCADD in the world (mostly from Europe), which were compiled from published case reports of each individual (29). The prevalence of clinical symptoms reported for patients in all three studies are summarized in Table 3. There is no data on genotype in these three studies; results are therefore for clinically or biochemically diagnosed MCADD and not specific to the G985A allelic variant. It is important to note that there is overlap between the studies referenced in Table 3. For example, some of the patients reported by Roe and Coates (1989) contribute to Iafolla et al's (1994) patient population, which include some patients from those reported from Touma and Charpentier (1992), which in turn include some patients from Vianey and Liaud's 1987 manuscript. However, since it is unknown how many or which patients are reported in duplicate among these studies, the studies are presented separately for the purposes of this review. Therefore, differences and similarities between the outcomes reported for each of these studies should be compared with the knowledge that there is some duplicate reporting among the studies. As shown in Table 3, hypoglycemia is the most common finding, present in over 90% of MCADD patients. The second most common symptoms are lethargy and vomiting. Most children with MCADD present with lethargy following a period of fasting, and more than half of MCADD patients exhibited vomiting. The study of Iafolla and colleagues was the only study to report encephalopathy. These findings are supported by the most recent prospective study on MCADD (6) in the U.K., where 37 of 46 MCADD patients (80%) suffered from acute hypoglycemia, and 23 of 46 (50%) experienced encephalopathy. The presence of seizures is less consistent among the three studies in Table 3, ranging from 17% to 43%. Despite the overwhelming incidence of coma in the first 23 MCADD patients reported (78%), coma was not assessed in the later studies; reasons for this include the possible reclassification as encephalopathy in later studies, or a possible decrease in coma due to interventions preventing coma associated with increased awareness of MCADD. Sudden death Two (22%) of 9 MCADD patients identified through newborn screening in Pennsylvania died suddenly and unexpectedly (5). In a study by the British Paediatric Association Surveillance System, 10 (21%) of 46 MCADD patients who presented acutely, died (6). Case studies have often documented sudden deaths among MCADD patients. The mean age of death among children with MCADD was 18.5 months. Among MCADD cases documented to date, patients have a 20-25% mortality rate at first presentation (27,30). However, these numbers may be lower if a number of MCADD patients remain undiagnosed during their lifetime. Furthermore, increased clinical awareness of fatty acid oxidation disorders in general may have significantly changed the observed outcomes over recent years. Although various studies have investigated the relationship between sudden death and MCADD, few have strictly investigated SIDS (defined as sudden and unexpected death among infants less than one year of age) and MCADD. These studies have varied in quality, and the results have been inconsistent. Furthermore, MCADD cases reported to date involve patients who are symptomatic even though many people with MCADD will be asymptomatic. Since the prevalence of symptoms is a risk factor for death, the relatively high incidence of death in clinically reported MCADD patients is not unexpected. Table 4 lists the estimated G985A carrier frequency in SIDS populations. These studies demonstrate that the frequency of this mutation in SIDS populations is similar to that in general newborn populations. These results indicate that this most common MCADD allelic variant has not accounted for much of the SIDS incidence in populations studied to date (12,13). Thus, universal newborn screening would not be expected to significantly lower rates of SIDS. Long-term outcomes Iafolla and colleagues (1994) have conducted the only long-term outcome evaluation of MCADD patients; their results are summarized in Table 5 (27). Developmental/behavioral disability. Developmental, behavioral, and neurological disabilities are beginning to be documented. Iafolla and colleagues (1994) followed 73 MCADD survivors more than 2 years of age; they found that 32% had abnormal results on formal psychodevelopmental tests, including speech and language delay and behavioral problems. Clinical onset between 12 and 18 months of age, which included seizures or encephalopathy, correlated strongly with the development of speech disability (27). A smaller study with longer follow-up reported handicaps and failure in school in MCADD patients (25). In Pollit and Leonard's 1998 study conducted with the British Pediatric Association Surveillance System, 6 of 36 surviving MCADD patients suffered neurological damage. Since follow-up has not been conducted past childhood or past the immediate post-illness period, it is unknown whether this number would increase if follow-up were performed through adulthood (40). Neurological defects among MCADD patients have also been documented in Denmark (41). Other long-term outcomes of MCADD include hypoglycemia and muscle weakness. Hypoglycemia can be induced by several measures such as illness or immunization. Treatment for hypoglycemia has been required for MCADD patients receiving the diphtheria-pertussis-tetanus vaccine, and for children with chicken-pox and otitis media. Risk factors for muscle weakness include an older age at diagnosis (>3 years of age), increased number of clinical events and hospitalizations before diagnosis, and time between clinical onset and diagnosis with longer delays increasing risk (27). Failure to thrive, seizure disorders, cerebral palsy, and ADD have also been reported as long-term outcomes among MCADD patients. A strong correlation existed between seizure at onset and cerebral palsy. Patients with ADD were more likely to have had seizures, encephalopathy, and hyperammonemia at the time of onset, as well as more episodes of clinical illness before and after diagnosis. MCADD patients diagnosed with ADD were also older at diagnosis of MCADD than patients who did not have ADD. The percentage of ADD cases attributable to MCADD is not known at present but is probably small. Interactions with Risk Factors MCADD results in only partial blockage of fatty acid oxidation; patients are therefore still able to metabolize short-chained fatty acids. As a result, a stimulus is needed for clinical symptoms to present. It is often in times of stress induced by fasting or infection when the demands on fatty acid oxidation are particularly high, that an MCADD patient may present with symptoms. Factors that affect clinical outcomes include lengthy fasting, the presence of infection or recent immunization, the patient's age when the first episode occurred, and a family history of SIDS or a sibling diagnosed with MCADD. In other words, the phenotype of MCADD is dependent on the degree of metabolic stress and a combination of other factors such as fever, infections, etc. (1). It is not known whether the phenotype of MCADD is also dependent on the allelic variant. Fasting Episodes of hypoketotic hypoglycemia, vomiting, lethargy, seizures and coma often are precipitated by fasting. Fasting leads to hypoglycemia, then vomiting or lethargy. Seizures and coma occur in fewer but more severe cases. The specific magnitude of risk associated with fasting has not yet been assessed or calculated. Infection/intercurrent illness Hypoglycemia requiring hospitalization is the most common outcome resulting from infection. This is often due to the vomiting and/or diarrhea brought on by the illness. Therefore, episodes of hypoketotic hypoglycemia, vomiting, lethargy, seizures and coma all occur in association with infection. If these episodes are severe, death can result. Three of 22 MCADD patients with chicken pox required treatment for hypoglycemia during an illness (27). One of two MCADD newborns who died in the supplemental testing program in Pennsylvania died during intercurrent illness (5). Immunizations Four percent of MCADD patients suffered from hypoglycemia requiring hospitalization after being immunized (27). One of the two MCADD patients identified through the supplemental newborn testing program in Pennsylvania died after a metabolic crisis stimulated by immunization (5). Age of first episode Disease severity varies with a patient's age of first presentation. Episodes usually occur when children are between 3 months and 3 years of age, with an average of 12 months. The risk of dying from MCADD is higher among patients whose first episode occurs after the age of one year. If the first episode occurs when the child is less than one year of age, the outcome usually consists of hypoglycemia and is not fatal. However, if the first episode doesn't occur until the child is greater than one year of age, then symptoms are "Reye-like" and can result in significant mortality. The highest risk for death occurs among children whose symptoms first appear when they are 15 to 26 months old (26). Roe and Coates (1989) reported that clinical onset between 12 and 18 months was usually marked by seizures or encephalopathy, and correlated strongly with the development of a speech disability. Twelve (19%) of 63 infants less than a year old died during their first MCADD episode. Nine (22%) of 41 children older than one year died during their first episode (26). Age of diagnosis The older a child is when the MCADD diagnosis is made, the higher the child's risk for muscle weakness and ADD (27). Risk for muscle weakness also increases as the time between clinical onset and diagnosis increases and the number of clinical events before diagnosis increases. Whether age is a true risk factor or simply a surrogate for delay in diagnosis needs to be explored. Seizures MCADD patients who present with seizures are at an increased risk for death, recurrent seizures, or cerebral palsy in the future (27). Family history A number of MCADD patients also had siblings who were affected with MCADD or who died unexpectedly. In half the families studied by Touma & Charpentier (1992), patients had one or more siblings who died prior to the proband's presentation (28). Of 43 families identified by Roe & Coates (1989), one-third had one or more unexplained sibling death (26). Iafolla and colleagues' (1994) study of 120 MCADD patients reported that 32% of patients had siblings who were either known to have MCADD or had died of SIDS. Of 57 cases collected prospectively by the British Paediatric Association Surveillance System between 1994 and 1996, 5 had siblings who died. Iafolla and colleagues (1991) reported that 20% of families (19 of 94) had one or more unexplained sibling death (42). Similarly, of 109 siblings of MCADD children, 8 had died from MCADD (23). Family history of MCADD is technically not a risk factor for having clinical symptoms; however, the association between the two is consistent with the autosomal recessive inheritance of the disease. A proportion of MCADD patients in all studies have siblings who are affected with MCADD or have siblings who died unexpectedly and suddenly, with some deaths diagnosed as SIDS. To date an average of 30% (10-50%) of MCADD patients have affected siblings. The presence of an affected sibling may thus be a risk factor for MCADD; it is also possible for the association to be due to ascertainment bias. Roe & Coates (1994) concluded that younger siblings of children with MCADD were the ones most at risk for a fatal first episode (23). However, the percentage of specific clinical outcomes among MCADD patients with affected siblings has not been determined. Studies comparing MCADD patients with family history of MCADD to MCADD patients identified in the population have not been conducted. Therefore, we do not know whether MCADD patients with affected siblings have a higher incidence of clinical symptoms than patients identified in the population. In other words, we do not know whether the penetrance of MCADD for clinical outcomes is higher among patients with a family history of the disorder than among patients identified in the general population. Table 6 lists the sensitivity, specificity, and positive predictive values of some of the main tests available for detecting people with MCADD. The main test currently used, MS/MS detects abnormal metabolites and is performed on either blood samples or urine. DNA/PCR testing can also be performed to confirm MCAD mutations in blood and dried-blood spots. Tandem mass spectrometry (MS/MS) is the most recent test developed and the most suitable for large-scale population screening for MCADD. MS/MS detects octanoylcarnitine in blood and is currently the method used in a supplemental screening program offered by Neo Gen Screening that screens most newborns in Pennsylvania and eastern Ohio. MS/MS is also the method of choice for North Carolina, California, and Massachusetts in their expanded newborn screening program that will include screening for MCADD. MS/MS screening is being piloted in the U.K. and Australia . North Carolina and Massachusetts are currently the only states testing for MCADD as part of their newborn screening program. Soon, California will offer optional testing for MCADD to parents. In addition, Neo Gen Screening offers voluntary MCADD testing to newborns born at birthing centers in the Northeast; 75% of Pennsylvania's newborns receive this test in addition to the testing provided by the Pennsylvania state screening program (21). In Europe, the Institute of Child Health in London is planning a population-based pilot study of MS/MS screening on newborns, which includes screening for MCADD. According to the 1997 British Health Technology Assessment report, MCADD satisfies most criteria for universal newborn screening. Its incidence in relevant populations is known, it is associated with significant morbidity or mortality, effective treatment is available, there is a period before onset of symptoms during which intervention improves outcome, and MS/MS allows safe and simple testing for the disease on the routine newborn screening blood spots. However, the first criterion for universal screening, that the natural history of the disorder be well-defined is not satisfied for MCADD (23). Furthermore, biochemical or clinical distinctions between allelic variants are also unknown. The lack of population data on MCADD leaves much of its clinical and natural history and long-term outcomes unknown. Therefore, the long-term benefits of MCADD screening have not been determined. Furthermore, there is no way of predicting which MCADD patients will be asymptomatic (23), therefore raising some concerns about population testing for MCADD. The most needed research on MCADD is therefore population-based studies that would demonstrate the utility of screening for MCADD. Internet Sites General Resources Genetic Databases Educational Resources Support Groups Related Slide Set
|
